Page 1 of 1

copying hash file

Posted: Sun Feb 05, 2006 7:21 pm
by htrisakti3
I need to copy hash files from one environment to another. I ftp'd the files: h_hashname & its directory, but when I tried to view data, it wouldn't open.

Thanks much for your help..

Re: copying hash file

Posted: Sun Feb 05, 2006 7:50 pm
by sanjay
Hello
u can't copy that way . u need write a job which will copy into sequential file and in u new environment 1 more job to copy back to hash file from sequential file.

Sanjay
htrisakti3 wrote:I need to copy hash files from one environment to another. I ftp'd the files: h_hashname & its directory, but when I tried to view data, it wouldn't open.

Thanks much for your help..

Posted: Sun Feb 05, 2006 8:04 pm
by htrisakti3
Unfortunately, these 2 environments are completely disconnected.

I need to copy the hash files from PROD which is secured, I am unable to even create jobs there. File systems is also secured which is off-limit for me.

I need to get copy of hash to diagnose when numbers don't reconcile. But I only have access to dev & test area.

Posted: Sun Feb 05, 2006 9:08 pm
by ray.wurlod
Tell us a bit more about the environments. For example is one using Intel chips and the other not, meaning that the byte order would need to be altered? Did you use a binary mode FTP?
A hashed file (note, "hashed", not "hash") consists of two file system objects that you have to move - a directory containing three files (DATA.30, OVER.30 and .Type30) and a file called D_HashedFileName which contains the metadata for the hashed file.
All of these need to be moved.
If you have done that successfully, using binary FTP, then View Data should work, provided your Hashed File stage uses "directory" as its access method.

Posted: Mon Feb 06, 2006 8:11 am
by I_Server_Whale
Ray is absolutely right. You should always use binary mode for moving hashed files. I did it recently and it works. But make sure that you move the hashed file directory along with its contents and also the D_hashed_file as Ray advised.

DATA.30 is the data file, OVER.30 is the overflow file and .Type30 indicates that the hashed file is a dynamic hashed file which is the default.

Also, kudos to Ray's patience. He's been correcting so many people that it is called a hashed and NOT a hash file. This has been the case with many posters and it has become ubiquitous.

If anyone comes across posters using 'hash' instead of 'hashed', please correct them instantly.

Posted: Mon Feb 06, 2006 8:39 am
by kumar_s
Also Job Sequence and not sequencer :wink:

-Kumar